Earticle

Automatic identification of residential areas from remote sensing images is beneficial to tasks such as urban planning and disaster assessment. Currently, residential area extraction tasks are primarily based on deep learning methods using single-modal data, and the information that a single modality can express is limited. Therefore, this paper proposes an end-to-end multi-modal semantic segmentation model that extracts features of remote sensing images and mobile phone signaling data through a dual-branch encoder, fuses the two features, and concatenates them with the feature map of the decoder stage. Experimental results show that our proposed method outperforms other models and can effectively identify residential areas.

With global concerns mounting over the aging of infrastructure and roadways, and the growing emphasis on road maintenance, safety inspections have become a critical priority. Current crack detection models face challenges related to feature loss and performance degradation. In response, we propose an Attention U-Net model designed to minimize information loss and improve crack detection performance, particularly in low-quality images. Additionally, we focus on optimizing image capture and detection on low-spec devices, leading to the development of a lightweight model that performs effectively even in resource-constrained environments.

In this study, a supervised deep learning approach is employed to construct a model that learns the relationship between input data and the corresponding output labels. The objective is to enable the model to generate accurate predictions for unseen data. Furthermore, we propose a method for semantically segmenting and tracking real-time drivable road areas to visually display them. To achieve this objective, the feasibility and performance of real-time applications using you only look once v8 (YOLOv8) architecture and mask regional-based convolutional neural network (Mask R-CNN) algorithm for learning are compared. Drivable road area segmentation involves delineating the present drivable zone across several lanes in the direction of vehicle travel, necessitating the detection of the status of the preceding vehicle. The identification rate for area detection is generally high during both day and night under clear conditions. Furthermore, the model derived from the YOLOv8 architecture demonstrated superior performance in visually extracting drivable area segmentation compared with the Mask R-CNN-based model.

Gabor filters well support computer vision tasks due to its steerable properties, that are useful for handling image transformation such as? image rotation and image scaling. In this paper, the Gabor-based UNet (GSAUNet) model is proposed where the Gabor filters replace the few default filters in the UNet. Besides, it has adopted a squeeze-and-excitation module to increase channel interdependencies and attention mechanism to focus on an important part of the network. The performance of GSAUNet is assessed on the CamVid and Cityscape datasets that attain improvement for the semantic segmentation tasks.

Current methods for semantic segmentation of remote-sensing images, especially for irregular and small targets, often result in low precision and incomplete feature extraction. To address this issue, an improved semantic segmentation method was developed utilizing DeepLabv3+. First, DeepLabv3+ is combined with the proposed feature fusion module to make full use of the complementary information of low- and high-level features. Second, the channel attention module helps extract effective features while suppressing irrelevant features, thereby enabling the extraction of more meaningful global information from high-level features. Finally, rich spatial information is selected using guided spatial attention, which improves the accuracy of edge segmentation of target objects. The results of the comparison show that the mean F1 score (MF1) and overall accuracy (OA) of the proposed method on the ISPRS Potsdam dataset are 89.81% and 88.45%, respectively. The MF1 of the proposed method is 89.90% and the OA is 89.14% for the UAVid dataset, which are higher than those of the other comparison algorithms. The proposed method exhibits superior semantic segmentation capabilities for remote-sensing images.

An image semantic segmentation model is proposed based on improved ENet network in order to achieve the low accuracy of image semantic segmentation in complex environment. Firstly, this paper performs pruning and convolution optimization operations on the ENet network. That is, the network structure is reasonably adjusted for better results in image segmentation by reducing the convolution operation in the decoder and proposing the bottleneck convolution structure. Squeeze-and-excitation (SE) module is then integrated into the optimized ENet network. Small-scale targets see improvement in segmentation accuracy via automatic learning of the importance of each feature channel. Finally, the experiment was verified on the public dataset. This method outperforms the existing comparison methods in mean pixel accuracy (MPA) and mean intersection over union (MIOU) values. And in a short running time, the accuracy of the segmentation and the efficiency of the operation are guaranteed.

This study examines the detection of drivable road areas using visual data from vehicle-mounted cameras, focusing on challenges such as curbed lanes and left and right turns. In this study, a modified DeepLabV3+ model, based on pixel-level semantic segmentation, is proposed to improve detection accuracy under varied driving conditions, with particular attention to visual representation. In the experiment, drivable road areas were segmented into four classes: background, straight road, left-turn road, and right-turn road. Performance comparisons between the proposed DeepLabV3+ model, which segmented each class into pixels, and the established Mask R-CNN algorithm showed that Mask R-CNN may achieve higher boundary accuracy but can reduce overall shape or connectivity. In contrast, DeepLabV3+ combines low-level spatial information with high-level semantic information, enabling precise pixel-level predictions. Thus DeepLabV3+ provides slightly lower boundary accuracy but produces more natural predictions of the overall road structure.

This paper is an attempt to design segmentation method based on fully convolutional networks (FCN) and attention mechanism. The first five layers of the Visual Geometry Group (VGG) 16 network serve as the coding part in the semantic segmentation network structure with the convolutional layer used to replace pooling to reduce loss of image feature extraction information. The up-sampling and deconvolution unit of the FCN is then used as the decoding part in the semantic segmentation network. In the deconvolution process, the skip structure is used to fuse different levels of information and the attention mechanism is incorporated to reduce accuracy loss. Finally, the segmentation results are obtained through pixel layer classification. The results show that our method outperforms the comparison methods in mean pixel accuracy (MPA) and mean intersection over union (MIOU).

Continual Test-Time Adaptation (CTTA) is essential for adapting models to target data in changing environments while retaining prior knowledge. However, previous methods overlook class imbalance, which limits performance on minor but essential objects. To address this, we propose AdaCoTTA, which applies confidence-guided adaptive learning to improve training stability. These mechanisms mitigate class imbalance, catastrophic forgetting, and error accumulation in CTTA. We evaluate AdaCoTTA on diverse CTTA benchmarks for semantic segmentation. It achieves state-of-the-art performance, improving average mIoU by 0.6% on ACDC and over 4% for minor classes, highlighting its effectiveness in scenarios. Code is available at?https://github.com/junghyeon0427/AdaCoTTA.

Deep convolution neural networks are the recent algorithms used for robotic vision. However, the complex crop?weed vegetation and the background interferences required a robust feature representation. Therefore, we proposed a Dual-branch Deep neural network for the semantic segmentation of crops and weeds. The branches utilized distinct feature extraction algorithms that extract essential semantic cues, and a decoder combined these features to improve the global contextual information. Finally, the hybrid feature selection module(HSFM) utilized the decoder features to complement one another. Experimental results show the proposed method obtained mean intersection of union scores of 0.8613 and 0.9099 on CWFID and BoniRob datasets, respectively.

With the development of artificial intelligence technology, various methods have been developed for recognizing objects in images using machine learning. Image segmentation is the most effective among these methods for recognizing objects within an image. Conventionally, image datasets of various classes are trained simultaneously. In situations where several classes require segmentation, all datasets have to be trained thoroughly. Such repeated training results in low training efficiency because most of the classes have already been trained. In addition, the number of classes that appear in the datasets affects training. Some classes appear in datasets in remarkably smaller numbers than others, and hence, the training errors will not be properly reflected when all the classes are trained simultaneously. Therefore, a new method that separates some classes from the dataset is proposed to improve efficiency during training. In addition, the accuracies of the conventional and proposed methods are compared.

차선이탈경고시스템(LDWS)은 첨단운전자지원시스템(ADAS)의 핵심 기능으로, 비의도적 차선 이탈을 탐지하여 사고를 예방한다. 그러나 기존의 에지 검출이나 허프 변환 기반 기법은 조명 변화, 차선 마모, 복잡한 도로 조건에서 한계가 있다. 본 연구는 딥러닝 기반 차선 세그멘테이션과 역원근 투영(IPM) 변환을 통합하여 LDWS 성능을 개선하는 방법을 제안한다. CamVid 데이터셋을 활용해 SegNet과 U-Net을 학습시켰으며, 결과를 탑뷰 영상으로 변환한 뒤 ResNet18 분류기를 통해 전처리 시나리오를 비교하였다. 실험 결과, U-Net이 SegNet보다 우수한 세그멘테이션 성능을 보였고, IPM 적용 시 분류 정확도가 크게 향상되었다. 특히 세그멘테이션 ROI와 IPM을 결합했을 때 최고 F1-score를 달성하여 균형 잡힌 검출 성능을 입증하였다. 따라서 제안 방법은 LDWS의 정확도와 강 인성을 향상시키는 효과적인 접근임을 확인하였다.

The Lane Departure Warning System (LDWS), a core function of Advanced Driver Assistance Systems (ADAS), prevents accidents by detecting unintended lane departures. Traditional methods based on edge detection or Hough transform are limited to illumination changes, lane fading, and complex roads. This study proposes integrating deep learning–based lane segmentation with Inverse Perspective Mapping (IPM) to improve LDWS performance. A SegNet and U-Net segmentation model was trained using the CamVid dataset, and the outputs were converted into top-view images, which were then evaluated with a ResNet18 classification model. The results show that U-Net achieved a higher level of segmentation accuracy, and IPM significantly improved classification accuracy. The combination of segmentation ROI and IPM yielded the best F1-score, confirming the effectiveness of the proposed approach for enhancing LDWS accuracy and robustness.

본 논문에서는 정밀한 semantic segmentation을 위해 강조 기법을 활용한 DeepLabv3+ 기반의 인코더-디코 더 모델을 제안하였다. DeepLabv3+는 딥러닝 기반 semantic segmentation 방법이며 자율주행 자동차, 적외선 이미 지 분석 등의 응용 분야에서 주로 사용된다. 기존 DeepLabv3+는 디코더 부분에서 인코더의 중간 특징맵 활용이 적어 복원 과정에서 손실이 발생한다. 이러한 복원 손실은 분할 정확도를 감소시키는 문제를 초래한다. 따라서 제안하는 방법 은 하나의 중간 특징맵을 추가로 활용하여 복원 손실을 최소화하였다. 또한, 추가 중간 특징맵을 효과적으로 활용하기 위해 작은 크기의 특징맵부터 계층적으로 융합하였다. 마지막으로, 디코더에 강조 기법을 적용하여 디코더의 중간 특징맵 융합 능력을 극대화하였다. 본 논문은 거리 영상 분할연구에 공통으로 사용되는 Cityscapes 데이터셋에서 제안하는 방법 을 평가하였다. 실험 결과는 제안하는 방법이 기존 DeepLabv3+와 비교하여 향상된 분할 결과를 보였다. 이를 통해 제안 하는 방법은 높은 정확도가 필요한 응용 분야에서 활용될 수 있다.

In this paper, we proposed a DeepLabv3+ based encoder-decoder model utilizing an attention mechanism for precise semantic segmentation. The DeepLabv3+ is a semantic segmentation method based on deep learning and is mainly used in applications such as autonomous vehicles, and infrared image analysis. In the conventional DeepLabv3+, there is little use of the encoder's intermediate feature map in the decoder part, resulting in loss in restoration process. Such restoration loss causes a problem of reducing segmentation accuracy. Therefore, the proposed method firstly minimized the restoration loss by additionally using one intermediate feature map. Furthermore, we fused hierarchically from small feature map in order to effectively utilize this. Finally, we applied an attention mechanism to the decoder to maximize the decoder's ability to converge intermediate feature maps. We evaluated the proposed method on the Cityscapes dataset, which is commonly used for street scene image segmentation research. Experiment results showed that our proposed method improved segmentation results compared to the conventional DeepLabv3+. The proposed method can be used in applications that require high accuracy.

본 논문에서는 U-Net 기반의 semantic segmentation 방법에서 정확도를 향상시키기 위해 residual learning을 활용한 인코더-디코더 구조의 모델을 제안하였다. U-Net은 딥러닝 기반의 semantic segmentation 방법 이며 자율주행 자동차, 의료 영상 분석과 같은 응용 분야에서 주로 사용된다. 기존 U-Net은 인코더의 얕은 구조로 인해 특징 압축 과정에서 손실이 발생한다. 특징 손실은 객체의 클래스 분류에 필요한 context 정보 부족을 초래하고 segmentation 정확도를 감소시키는 문제가 있다. 이를 개선하기 위해 제안하는 방법은 기존 U-Net에 특징 손실과 기울기 소실 문제를 방지하는데 효과적인 residual learning을 활용한 인코더를 통해 context 정보를 효율적으로 추출 하였다. 또한, 인코더에서 down-sampling 연산을 줄여 특징맵에 포함된 공간 정보의 손실을 개선하였다. 제안하는 방법은 Cityscapes 데이터셋 실험에서 기존 U-Net 방법에 비해 segmentation 결과가 약 12% 향상되었다.

In this paper, we proposed an encoder-decoder model utilizing residual learning to improve the accuracy of the U-Net-based semantic segmentation method. U-Net is a deep learning-based semantic segmentation method and is mainly used in applications such as autonomous vehicles and medical image analysis. The conventional U-Net occurs loss in feature compression process due to the shallow structure of the encoder. The loss of features causes a lack of context information necessary for classifying objects and has a problem of reducing segmentation accuracy. To improve this, The proposed method efficiently extracted context information through an encoder using residual learning, which is effective in preventing feature loss and gradient vanishing problems in the conventional U-Net. Furthermore, we reduced down-sampling operations in the encoder to reduce the loss of spatial information included in the feature maps. The proposed method showed an improved segmentation result of about 12% compared to the conventional U-Net in the Cityscapes dataset experiment.

본 논문에서는 U-Net 기반의 semantic segmentation 방법에서 정확도를 개선하기 위한 Atrous Residual U-Net (AR-UNet)을 제안하였다. U-Net은 의료 영상 분석, 자율주행 자동차, 원격 감지 영상 등의 분야에서 주로 사용 된다. 기존 U-Net은 인코더 부분에서 컨볼루션 계층 수가 적어 추출되는 특징이 부족하다. 추출된 특징은 객체의 범주를 분류하는 데 필수적이며, 부족할 경우 분할 정확도를 저하시키는 문제를 초래한다. 따라서 이 문제를 개선하기 위해 인코 더에 residual learning과 ASPP를 활용한 AR-UNet을 제안하였다. Residual learning은 특징 추출 능력을 개선하고, 연속적인 컨볼루션으로 발생하는 특징 손실과 기울기 소실 문제 방지에 효과적이다. 또한 ASPP는 특징맵의 해상도를 줄이지 않고 추가적인 특징 추출이 가능하다. 실험은 Cityscapes 데이터셋으로 AR-UNet의 효과를 검증하였다. 실험 결과는 AR-UNet이 기존 U-Net과 비교하여 향상된 분할 결과를 보였다. 이를 통해 AR-UNet은 정확도가 중요한 여러 응용 분야의 발전에 기여할 수 있다.

In this paper, we proposed an Atrous Residual U-Net (AR-UNet) to improve the segmentation accuracy of semantic segmentation method based on U-Net. The U-Net is mainly used in fields such as medical image analysis, autonomous vehicles, and remote sensing images. The conventional U-Net lacks extracted features due to the small number of convolution layers in the encoder part. The extracted features are essential for classifying object categories, and if they are insufficient, it causes a problem of lowering the segmentation accuracy. Therefore, to improve this problem, we proposed the AR-UNet using residual learning and ASPP in the encoder. Residual learning improves feature extraction ability and is effective in preventing feature loss and vanishing gradient problems caused by continuous convolutions. In addition, ASPP enables additional feature extraction without reducing the resolution of the feature map. Experiments verified the effectiveness of the AR-UNet with Cityscapes dataset. The experimental results showed that the AR-UNet showed improved segmentation results compared to the conventional U-Net. In this way, AR-UNet can contribute to the advancement of many applications where accuracy is important.

In this project, we present SpudScan, a comprehensive approach for the semantic segmentation of potato leaf diseases. Utilizing a custom dataset collected and meticulously annotated by our team, we trained and evaluated three state-of-the-art models: Segformer, Unet, and Deeplab V3. Our objective was to accurately identify and segment various disease-affected regions on potato leaves to facilitate early detection and treatment. Comparative analysis of the models highlights their respective strengths and weaknesses in terms of performance, processing time, and robustness. The results demonstrate significant potential for integrating deep learning techniques in agricultural disease management, paving the way for more efficient crop monitoring and health assessment.